Assessment of Red Sea temperatures in CMIP5 models for present and future climate

10˚C�day in the open Indian Ocean (Fig 10G). This spatial distribution is mainly determined by the length of the MHWs (see Fig 10A). For the end of the century, the models project a large increase in the AAMI with a spatial pattern very similar under both scenarios, with the Fig 10. Statistics of MHWs for present conditions (left column), scenario RCP8.5 (ensemble average, middle column) and associated uncertainties (ensemble STD, right column). (a-c) Average length of MHWs in days per year. (d-f) Annual mean Intensity (˚C). (g-i) Annual Accumulative mean Intensity (˚C � day). Note the colorbars for the future scenarios are different than for the present conditions. https://doi.org/10.1371/journal.pone.0255505.g010 PLOS ONE Assessment of Red Sea temperatures in CMIP5 models for present and future climate PLOS ONE | https://doi.org/10.1371/journal.pone.0255505 July 30, 2021 17 / 26

largest values in the southern Red Sea and the lower values in the Outer Region. In particular AAMI would reach about 150 ± 35˚C�day under scenario RCP4.5, with maximum values in the southern Red Sea (180 ± 55˚C�day), decreasing to 50 ± 30˚C�day approximately in the Outer Region (S9H and S9I Fig). Under scenario RCP8.5 the projected AAMI is 300– 400 ± 100˚C�day in the northern Red Sea and around 250 ± 120˚C�day in the Outer Region (Fig 10H and 10I). Discussion GCMs are not designed to simulate regional processes, mainly because of their coarse resolu- tion and the lack of specific tuning of model parameters, which is the main reason why GCMs are often discarded for regional analyses in semi enclosed seas [39]. Previous studies compar- ing regional and global models with observations suggest that regional simulations are required to properly simulate key processes in semi enclosed seas [16,40]. That would be the case for those analyses focused on the circulation and the 3D heat and salt distribution, which require a minimum spatial resolution to be properly solved. However, [18] showed that a rela- tively simple process like the air-sea interaction, can be reasonably reproduced by GCMs. Therefore, ocean variability which would only depend on the air-sea interaction could poten- tially be faithfully represented by GCMs. In this sense, [18] reported that projected changes of surface temperature in the Mediterranean for the 21st century were similar using an ensemble of GCMs or regional climate models. In conclusion, the suitability of GCMs for a particular

study is case dependent. In order to further investigate the driving mechanism for the Red Sea heat content variability we compare observed heat content in the upper 100 m in the Red Sea from the TEMPERSEA product [17] with air temperatures at 1000 mbars obtained from the JRA55 reanalysis [16], (https://rda.ucar.edu/datasets/ds628.1/?hash=access). The comparison of the yearly values for the period 1958–2013 (Fig 11) shows that interannual variations in the Red Sea heat content are highly correlated to variations in the air temperature (r = 0.68). It has to be noted that both datasets are totally independent as one is constructed from ocean obser- vations and the other using atmospheric data. This result reinforces the idea that heat content variations could be reproduced, at least as a first approximation, by models correctly reproduc- ing air-sea interactions. In order to assess if GCMs are able to reproduce the heat content variability in the Red Sea, here we have assessed the skills of 43 GCMs in terms of different diagnostics related to heat uptake. Our results are very heterogeneous with some of the models that did not even repre- sent the Red Sea in their ocean mask, and others that showed very poor results in all the diag- nostics. However, we have also found that some GCMs were able to reproduce the present variability with an accuracy comparable to the reference regional simulations (KAUST and GLORYS), especially inside the Red Sea. Those models are able to approximate reasonably

well the seasonal and interannual variability of the temperature profile in the upper 300 m, suggesting that ocean heat uptake is properly modelled. Also, they are able to represent the spatial variability with the north to south gradients in the Red Sea and the distinctive behav- iour of the Outer Region. We found the skill of GCMs to be typically better inside the Red Sea than in the Outer zone. This is consistent with a previous study [17], that has shown that the heat content evolution in the Red Sea basin is mainly driven by the air-sea interactions, while the Outer Region is also affected by the advection of waters from the open Indian ocean. This suggests that the mecha- nisms underpinning the basin temperature evolution of the Red Sea are relatively simpler, and thereby easier to be captured by GCMs. In the Outer Region, GCMs should be able to properly model not only the air-sea interactions but also the temperature evolution and the circulation PLOS ONE Assessment of Red Sea temperatures in CMIP5 models for present and future climate PLOS ONE | https://doi.org/10.1371/journal.pone.0255505 July 30, 2021 18 / 26

patterns in the Indian Ocean. Thus, in our case, the isolation of the Red Sea from open ocean dynamics makes easier for the GCM to represent the heat content variability. The temperature projections have been done based on the ensemble of 8 models that have met acceptable levels in selected diagnostics. Those diagnostics have been selected to reach a balance between the number of requirements and the number of models that are needed for the computations. However, as we have mentioned above, there is an unavoidable degree of subjectivity in any model selection procedure. Therefore, for completeness we have computed also the projected SST anomaly for different subensembles created following different criteria (Table 1). The results show that depending on the criterion chosen the projected SST increase under scenario RCP8.5 would range between 3.15˚C and 3.45˚C (S10 Fig). The former is obtained using criterion 1, the most restrictive, in which only 2 models complied (see Table 2). The latter was obtained using criterion 4, but in this case, there were no requirements on the GCM trends (Table 2). Namely, models with present climate trends larger than the observed range (including uncertainties) are kept and drive the ensemble towards larger projected anomalies. Therefore, the selected criterion (number 2), seems to provide a good compromise, between realistic representation of key processes and the number of models kept in the ensem- ble. Nevertheless, the final results would not have changed significantly if other criterion

to be considered as a first approximation. Higher resolution models would be able to repro- duce more mechanisms and to provide more accurate projections. However, until an ensemble of high resolution ocean models is available the projections done by the selected CMIP5 GCMs can be considered as a benchmark result. A recent study by [41], identified a multidecadal variability of Red Sea SST with a period of 70 years and an amplitude of approximately 0.3˚C. Those authors suggested the Red Sea could enter into a cooling phase in the next years. However, that conclusion was derived from a rough extrapolation of the observations and can be reviewed in the light of the outcome of the more sophisticated approach presented here (S11 Fig). First, the observed multidecadal oscilla- tion is overimposed to a long-term trend, so both components should be examined jointly. Depending on the long-term evolution during the next years, the absolute change will have a sign or another. Second, the concept of cooling requires the definition of a reference period. Overimposing the 70 years oscillation identified by [41] to the 20th century trend, the Red Sea SST would reach a relative minimum of 27.6˚C in 2035–2040, 0.2˚C lower than the value recorded during the 2000s but still higher than the mean value of the 20th century (27.5˚C). However, this ignores global warming. If the 70 years oscillation is superimposed on the trend observed during the satellite period (1985–2019), the absolute temperatures would remain

roughly constant till year 2035–2040 and then rise at an even higher rate due to the entering into the positive phase of the multidecadal oscillation. Finally, considering warming under “business as usual” RCP8.5 projected trends, the Red Sea SST will warm continuously even in the presence of a 70 years oscillation (S11 Fig). The characterization of marine heatwaves is also of great importance for the Red Sea, where several episodes of coral bleaching have been reported in the past [29]. Due to the limited avail- ability of high-frequency data for the selected GCMs, we have focused on the analysis of how the mean temperature change would affect the MHW characteristics. To do that we have pro- jected the daily SST assuming that intraseasonal variability remains constant during the next decades (i.e. adding the projected monthly temperature anomaly to observed daily values). This has proven to be a reasonable assumption for the Mediterranean Sea, where only a 10– 20% of the changes in MHW characteristics can be attributed to changes in the day-to-day var- iability [18,22]. Moreover, at global scale, [23] analysed the high frequency outputs from an ensemble of GCMs and reached the conclusion that the changes in the occurrence of MHWs are mainly driven by the global-scale shift in mean SSTs. This extent has also been confirmed by the observational study of [7,8]. Our study predicts a higher intensity of MHWs in the southern part of the Red Sea (Fig 10F and 10I), even if the warming pattern is similar throughout the region and the intraseasonal

variability remains unchanged (see Supplementary Information). The reason for the sensitivity of the southern Red Sea to MHWs is found in the shape of the seasonal cycle, which is flatter in that region, resulting in an extended summer season. Therefore, once the mean summer temperature exceeds the local threshold, there would be many days per year exceeding that threshold. In the Northern Region or even more in the Outer Region, the summer is much shorter, so the increase in the mean value would have a more moderate impact on the MHW characteristics (Fig 12). Finally, our results highlight the benefits of mitigation. The averaged warming in the sea surface would go from 3.3±0.6˚C under a business as usual scenario to 1.6±0.4˚C under a sce- nario of moderate mitigation (RCP4.5), with respect to period 1990–2010. Additionally, we can do a first assessment of what would be the situation in the Red Sea if the goals of the Paris agreement were reached and the global temperature increase with respect to preindustrial period remained below 1.5˚C. That situation would happen under scenario RCP8.5 between years 2025–2040 (most likely 2038) and under scenario RCP4.5 between years 2040–2075 PLOS ONE Assessment of Red Sea temperatures in CMIP5 models for present and future climate PLOS ONE | https://doi.org/10.1371/journal.pone.0255505 July 30, 2021 20 / 26

(most likely 2045). In those periods and scenarios the SST increase in the Red Sea would be of only 0.7 ± 0.6˚C, which also matches the results obtained with 3 GCMs under scenario RCP2.6 by [13]. Therefore, it is clear that the expected impacts in the Red Sea in terms of mean warm- ing but also in terms of MHWs would be much reduced under a scenario of strong mitigation. Conclusions In this work we have assessed the skills of coarse resolution GCMs in reproducing the heat content variablity of the Red Sea. Namely we used different diagnostics on the seasonal, inter- annual and long-term variability of the temperature field. The results have shown very diverse results for the different GCMs, with almost half of them being unacceptable under the pro- posed criteria. However, we have also found a limited subensemble of 8 GCM models that show acceptable skills in most of the diagnostics, at a level comparable to two reference high resolution simulations. The selected subensemble was then used to project the temperature evolution in the Red Sea and the Gulf of Aden under a business-as-usual scenario (RCP8.5) and a moderate scenario (RCP4.5). The models project a larger warming under the former (3.3 ±0.6˚C at the surface for the period 2080–2100) than under the latter (1.6 ±0.4˚C at the surface for the period 2080–2100). Also, the models consistently showed that under scenario RCP8.5 the water column would become more stratified, while under scenario RCP4.5 the warming is expected to be quite homogenous in depth. This rise of the mean temperature

would also affect the characteristics of marine heat waves. At present, the number of days with Fig 12. Daily SST averaged in the north (Top panel), south (middle panel) and Outer Region (bottom panel) for present conditions (grey), under scenario RCP 4.5 (light blue) and under scenario RCP8.5 (orange). The dots indicate the occurrence of MHWs in present conditions (black), under scenario RCP 4.5 (blue) and under scenario RCP 8.5 (red). For clarity only a limited period is plotted (1998–2003 for the present conditions, bottom axis and 2090–2096 for the future conditions, top axis). https://doi.org/10.1371/journal.pone.0255505.g012 PLOS ONE Assessment of Red Sea temperatures in CMIP5 models for present and future climate PLOS ONE | https://doi.org/10.1371/journal.pone.0255505 July 30, 2021 21 / 26

MHW conditions ranges among 8 and 12 days per year in all the region. At the end of the cen- tury, under the RCP4.5 scenario, the number of days would increase to 125 days per year in the southern Red Sea and the Gulf of Aden, and around 100 days per year in the rest of the region. Under the RCP 8.5 scenario the situation would be even worse, reaching more than 200 days per year in the southern Red Sea and around 150 days in the northern part and the Gulf of Aden. Consequently, the annual mean intensity (0.6˚C) of the MHWs would increase in the Red Sea, being around 4 (3.6) times larger under the scenario RCP8.5 (RCP4.5). These projections based on GCMs have to be considered as a first approximation and a benchmark result until an ensemble of high resolution ocean models is available. Nevertheless, they clearly point towards the danger of a business-as-usual scenario of emissions and the great potential benefits that mitigation would bring to the Red Sea. Supporting information S1 Fig. Same as Fig 3 but for the outer region. KAUST simulation does not cover this area and is therefore not included in the plots. (TIF) S2 Fig. Vertical profile of the interannual STD (in ˚C) in the Red Sea (left) and in the Gulf of Aden (right) for observations (black line), KAUST (red line), GLORYS (blue line), CMIP5 models (grey lines) and the average of selected GCMs (see “Model Selection” section, green line). (TIF) S3 Fig. Maps of averaged SST anomaly for the period 1985–2005 (in ˚C) with respect to the basin average for each selected GCMs. (TIF)

S4 Fig. Left panel: Vertical profile of averaged temperatures in the Red Sea (in ˚C) for the observations (black), RCP4.5 (blue dashed lines, all models selected) and RCP8.5 (red dashed lines, all models selected) and the ensemble average is represented with a thick line. Right panel: Vertical profile of projected anomalies in the Red Sea for all the selected models under scenario RCP4.5 (blue) and RCP8.5(red). The ensemble average is represented with a thick line. (TIF) S5 Fig. (Top) Anomaly of the observed Sea Surface temperature (blue) and fitted seasonal cycle (red). (Bottom) Intra-seasonal variability of the observed sea surface temperature (grey). (TIF) S6 Fig. Seasonal climatology of the selected CMIP5 models for Present (grey lines) and Future climate (red lines). The values of each model are represented with thin lines, while the ensemble average is represented with thick lines. (TIF) S7 Fig. Simulated changes in different MHWs descriptors presented as the ratio of pro- jected values over the present climate values. (a) annual mean duration, (b) annual mean intensity and (c) annual cumulative mean intensity (d). The colours represent different thresh- olds that have been used to define a MHW (90th, 95th, 98th and 99th percentiles). (TIF) S8 Fig. Marine Heat Wave (MHW) characteristics in the present climate. 95th -Percentile (left panel) in ˚C, number of MHWs per year (middle panel), and the annual mean intensity PLOS ONE Assessment of Red Sea temperatures in CMIP5 models for present and future climate

PLOS ONE | https://doi.org/10.1371/journal.pone.0255505 July 30, 2021 22 / 26

(AMI) in ˚C (right panel). (TIF) S9 Fig. Statistics of MHWs for present conditions (left column), scenario RCP4.5 (ensemble average, middle column) and associated uncertainties (ensemble STD, right column). (a-c) Average length of MHWs in days per year. (d-f) Annual mean Intensity (˚C). (g-i) Annual Accumulative mean Intensity (˚C � day). Note the colorbars for the future scenarios are differ- ent than for the present conditions. (TIF) S10 Fig. Projected SST anomaly for different subensembles according to different criteria (see Table 1). (TIF) S11 Fig. Simplified projections for the evolution of basin averaged SST. Assuming the mul- tidecadal oscillation identified by Krokos et al., (2019) continues along the 21st century, we test how the temperatures would change imposing different long term trends: The observed 20st century trend (in black), the trend observed during the satellite period (yellow) or the trend projected by models under scenario RCP8.5 (red). (TIF) S1 Table. Set of diagnostics to rank the performance of the models from the model outputs and the observational datasets for the period 1985–2005, for the Red Sea and the Gulf of Aden (Outer Region) separately. (XLSX) S2 Table. Set of diagnostics to rank the performance of the models from the model outputs and the observational datasets for the period 1985–2005, for the Gulf of Aden (Outer Region). (XLSX) S3 Table. Change of the basin averaged sea surface temperature by the end of the century (2080–2100) respect to the present (1985–2005) of those GCM that fulfil at least one of the criterion (see Table 1).

(XLSX) S4 Table. Statistics of MHWs for each selected GCM. Average length of MHWs (DMHW, in days/year), Annual Mean Intensity (AMI, in ˚C/year) and Annual Acumulative Mean Intensity (AAMI, in ˚C x day/year). Results are presented for the present climate (1985–2015) and for projected future conditions under scenarios RCP4.5 and RCP8.5 for the period 2080– 2100. (XLSX) Acknowledgments We thank marine biologists and physical oceanographers who have collected data and ana- lysed the Red Sea mechanisms for the last decades. We specially acknowledge the support of the King Abdullah University of Science and Technology (KAUST), for providing data and knowledge about the temperature evolution and biology interactions of the Red Sea. We also thank Mediterranean Institute for Advanced Studies (IMEDEA, UIB/CSIC) and the Spanish Institute of Oceanography (IEO/CSIC) for providing the computational resources to address the goals of this work. PLOS ONE Assessment of Red Sea temperatures in CMIP5 models for present and future climate PLOS ONE | https://doi.org/10.1371/journal.pone.0255505 July 30, 2021 23 / 26

Author Contributions Conceptualization: Gabriel Jordà, Carlos M. Duarte. Data curation: Miguel Agulles, Ibrahim Hoteit. Formal analysis: Miguel Agulles. Funding acquisition: Carlos M. Duarte. Investigation: Miguel Agulles, Gabriel Jordà, Carlos M. Duarte. Methodology: Miguel Agulles, Gabriel Jordà. Resources: Carlos M. Duarte. Supervision: Gabriel Jordà. Validation: Miguel Agulles. Writing – original draft: Miguel Agulles. Writing – review & editing: Gabriel Jordà, Ibrahim Hoteit, Susana Agustı´, Carlos M. Duarte. References 1. Thorne K. S., Wheeler J. A., Francisco S., Wieman C. E., Ertmer W., Schleich W. P. et al: References and Notes 1., 327(June), 1543–1548, https://doi.org/10.1126/science.1206034, 2010. PMID: 22053051 2. Hughes T. P., Anderson K. D., Connolly S. R., Heron S. F., Kerry J. T., Lough J. et al: Spatial and tem- poral patterns of mass bleaching of corals in the Anthropocene, Science (80-.)., 359(6371), 80–83, https://doi.org/10.1126/science.aan8048, 2018. PMID: 29302011 3. Monroe A. A., Ziegler M., Roik A., Ro¨ thig T., Hardenstine R. S., Emms M. A., et al: In situ observations of coral bleaching in the central Saudi Arabian Red Sea during the 2015/2016 global coral bleaching event, PLoS One, 13(4), 1–13, https://doi.org/10.1371/journal.pone.0195814, 2018. PMID: 29672556 4. Osman E. O., Smith D. J., Ziegler M., Ku¨ rten B., Conrad C., El-Haddad K. M., et al: Thermal refugia against coral bleaching throughout the northern Red Sea, Glob. Chang. Biol., 24(2), e474–e484, https://doi.org/10.1111/gcb.13895, 2018. PMID: 29044761

5. Poloczanska E. S., Burrows M. T., Brown C. J., Garcı´a Molinos J., Halpern B. S., Hoegh-Guldberg O. et al: Responses of Marine Organisms to Climate Change across Oceans, Front. Mar. Sci., 3(May), 1– 21, https://doi.org/10.3389/fmars.2016.00062, 2016. 6. Gittings et al.: Mass spawning and reproductive viability of reef corals at the East Flower Garden Bank, northwest Gulf of Mexico, Bull. Mar. Sci., 51(3), 420–428, 1992. 7. Oliver E. C. J.: Mean warming not variability drives marine heatwave trends, Clim. Dyn., 53(3–4), 1653–1659, 2019. https://doi.org/10.1016/j.scitotenv.2019.135992 PMID: 31841928 8. Oliver E. C. J., Donat M. G., Burrows M. T., Moore P. J., Smale D. A., Alexander L. V, et al: Longer and more frequent marine heatwaves over the past century, Nat. Commun., 9(1), 1–12, https://doi.org/10. 1038/s41467-017-02088-w, 2018. PMID: 29317637 9. Chaidez V., Dreano D., Agusti S., Duarte C. M. and Hoteit I.: Decadal trends in Red Sea maximum sur- face temperature, Sci. Rep., 7(1), 1–8, https://doi.org/10.1038/s41598-016-0028-x, 2017. PMID: 28127051 10. Burrows M. T., Schoeman D. S., Buckley L. B., Moore P., Poloczanska E. S., Brander K. M., et al: The pace of shifting climate in marine and terrestrial ecosystems, Science (80-.)., 334(6056), 652–655, https://doi.org/10.1126/science.1210288, 2011. PMID: 22053045 11. Zenetos A.; C¸ inar M.E.; Pancucci-Papadopoulou M.A.; Harmelin J.G.; Furnari G.; Andaloro F., et al; and Zibrowius H.: Annotated list of marine alien species in the Mediterranean with records ot the worst invasive species, 4, 57–66, 2005.

12. Cantin N. E., Cohen A. L., Karnauskas K. B., Tarrant A. M., & McCorkle D. C. (2010). Ocean warming slows coral growth in the central Red Sea. Science, 329(5989), 322–325. https://doi.org/10.1126/ science.1190182 PMID: 20647466 13. Shaltout M.: Recent sea surface temperature trends and future scenarios for the Red Sea, Oceanolo- gia, 61(4), 484–504, https://doi.org/10.1016/j.oceano.2019.05.002, 2019. PLOS ONE Assessment of Red Sea temperatures in CMIP5 models for present and future climate PLOS ONE | https://doi.org/10.1371/journal.pone.0255505 July 30, 2021 24 / 26

14. Taylor K. E., Stouffer R. J. and Meehl G. A.: An Overview of CMIP5 and Experimental Design, Bull. Am. Meteorol. Soc., 93(april), 485–498, https://doi.org/10.1175/BAMS-D-11-00094, 2012. 15. Jacob D., Petersen J., Eggert B., Alias A., Christensen O. B., Bouwer L. M., et al.: EURO-CORDEX: New high-resolution climate change projections for European impact research, Reg. Environ. Chang., 14(2), 563–578, https://doi.org/10.1007/s10113-013-0499-2, 2014. 16. Calafat F. M., Jorda´ G., Marcos M. and Gomis D.: Comparison of Mediterranean sea level variability as given by three baroclinic models, J. Geophys. Res. Ocean., 117(C2), 2012. 17. Agulles M., Jordà G., Jones B., Agustı´ S. and Duarte M. C.: Temporal evolution of temperatures in the red sea and the gulf of aden based on in situ observations (1958–2017), Ocean Sci., 16(1), https://doi. org/10.5194/os-16-149-2020, 2020. 18. Jordà G., Marbà N. and Duarte C. M.: Mediterranean seagrass vulnerable to regional climate warming, Nat. Clim. Chang., 2(11), 821–824, https://doi.org/10.1038/nclimate1533, 2012. 19. Somot S., Ruti P., Ahrens B., Coppola E., Jordà G., Sannino G. and Solmon F.: Editorial for the Med- CORDEX special issue, Clim. Dyn., 51(3), 771–777, https://doi.org/10.1007/s00382-018-4325-x, 2018. 20. Parent L. et al.: GLORYS ¼ ˚ Global Ocean Reanalysis and Simulations of the Period 1992-Present, 1992, 2011. 21. Fengchao Yao; Ibrahim Hoteit; Larry J.Pratt; Amy S.Bower; Ping Zhai; Armin Ko¨ hl; and Ganesh Gopa- lakrishnan: Seasonal overturning circulation in the Red Sea: 1. Model validation and summer circula-

tion, in Journal of Geophysical Research: Oceans, vol. 119, pp. 1365–1382., 2014. 22. Darmaraki S., Somot S., Sevault F., Nabat P., Cabos Narvaez W. D., Cavicchia L., et al.: Future evolu- tion of Marine Heatwaves in the Mediterranean Sea, Clim. Dyn., 0(0), 0, https://doi.org/10.1007/ s00382-019-04661-z, 2019. 23. Fro¨ licher T. L., Fischer E. M. and Gruber N.: Marine heatwaves under global warming, Nature, 560 (7718), 360–364, https://doi.org/10.1038/s41586-018-0383-9, 2018. PMID: 30111788 24. Hobday A. J., Alexander L. V., Perkins S. E., Smale D. A., Straub S. C., Oliver E. C. J., et al: A hierarchi- cal approach to defining marine heatwaves, Prog. Oceanogr., 141, 227–238, https://doi.org/10.1016/j. pocean.2015.12.014, 2016. 25. Wernberg T., Smale D. A., Tuya F., Thomsen M. S., Langlois T. J., De Bettignies T., et al.: An extreme climatic event alters marine ecosystem structure in a global biodiversity hotspot, Nat. Clim. Chang., 3 (1), 78–82, https://doi.org/10.1038/nclimate1627, 2013. 26. Garrabou J., Coma R., Bensoussan N., Bally M., Chevaldonne´ P., Cigliano M., et al: Mass mortality in Northwestern Mediterranean rocky benthic communities: effects of the 2003 heat wave, Glob. Chang. Biol., 15(5), 1090–1103, 2009. 27. Marbà N. and Duarte C. M.: Mediterranean warming triggers seagrass (Posidonia oceanica) shoot mor- tality, Glob. Chang. Biol., 16(8), 2366–2375, 2010. 28. Genevier L. G. C., Jamil T., Raitsos D. E., Krokos G. and Hoteit I.: Marine heatwaves reveal coral reef

zones susceptible to bleaching in the Red Sea, Glob. Chang. Biol., (April), https://doi.org/10.1111/gcb. 14652, 2019. PMID: 30974020 29. Harada Y., Ebita A., Ota Y., Onogi K., Takahashi K., Endo H., et al: The Jra-55 Reanalysis: General Specifications And Basic Characteristics, J. Meteorol. Soc. Japan. Ser. Ii, 93(1), 5–48, https://doi.org/ 10.2151/Jmsj.2015-001, 2015. 30. Thomson A. M., Calvin K. V., Smith S. J., Kyle G. P., Volke A., Patel P., et al. A.: RCP4.5: A pathway for stabilization of radiative forcing by 2100, Clim. Change, 109(1), 77–94, https://doi.org/10.1007/s10584- 011-0151-4, 2011. 31. Marshall J., Hill C., Perelman L. and Adcroft A.: Hydrostatic, quasi-hydrostatic, and nonhydrostatic ocean modeling, 102, 5733–5752, 1997. 32. Dee D. P., Uppala S. M., Simmons A. J., Berrisford P., Poli P., Kobayashi S., et al: The ERA-Interim reanalysis: Configuration and performance of the data assimilation system, Q. J. R. Meteorol. Soc., 137(656), 553–597, https://doi.org/10.1002/qj.828, 2011. 33. Zhan P., Subramanian A. C., Yao F., Kartadikaria R., A., Guo D. and Hoteit I.: The eddy kinetic energy budget in the Red Sea, J. Geophys. Res. Ocean., 121(5), 3010–3028, https://doi.org/10.1002/ 2015JC011589, 2016. 34. Dreano: The Gulf of Aden Intermediate Water Intrusion Regulates the Southern Red Sea Summer Phy- toplankton Blooms, Oceanologia, 49(3), 337–351, https://doi.org/10.1371/journal.pone.0168440, 2016. PMID: 28006006 35. Gittings J. A., Raitsos D. E., Krokos G. and Hoteit I.: Impacts of warming on phytoplankton abundance

and phenology in a typical tropical marine ecosystem, Sci. Rep., 8(1), 1–12, https://doi.org/10.1038/ s41598-017-17765-5, 2018. PMID: 29311619 PLOS ONE Assessment of Red Sea temperatures in CMIP5 models for present and future climate PLOS ONE | https://doi.org/10.1371/journal.pone.0255505 July 30, 2021 25 / 26

36. Raitsos D. E., Brewin R. J. W., Zhan P., Dreano D., Pradhan Y., Nanninga G. B. et al: Sensing coral reef connectivity pathways from space, Sci. Rep., 7(1), 1–10, https://doi.org/10.1038/s41598-016- 0028-x, 2017. PMID: 28127051 37. Cabanes C., Grouazel A., Von Schuckmann K., Hamon M., Turpin V., Coatanoan, et al: The CORA dataset: Validation and diagnostics of in-situ ocean temperature and salinity measurements, Ocean Sci., 9(1), 1–18, https://doi.org/10.5194/os-9-1-2013, 2013. 38. Hayashida H., Matear R. J., Strutton P. G. and Zhang X.: Insights into projected changes in marine heatwaves from a high-resolution ocean circulation model, Nat. Commun., 11(1), 1–9, https://doi.org/ 10.1038/s41467-020-18241-x, 2020. PMID: 32859903 39. Llasses J., Jordà G., Gomis D., Adloff F., Macı´as D., Harzallah A., et al.: Heat and salt redistribution within the Mediterranean Sea in the Med-CORDEX model ensemble, Clim. Dyn., 51(3), 1119–1143, https://doi.org/10.1007/s00382-016-3242-0, 2018. 40. Giorgi F. and Lionello P.: Climate change projections for the Mediterranean region, Glob. Planet. Change, 63(2–3), 90–104, https://doi.org/10.1016/j.gloplacha.2007.09.005, 2008. 41. Krokos G., Papadopoulos V. P., Sofianos S. S., Ombao H., Dybczak P. and Hoteit I.: Natural Climate Oscillations may Counteract Red Sea Warming Over the Coming Decades, Geophys. Res. Lett., 46 (6), 3454–3461, https://doi.org/10.1029/2018GL081397, 2019. PLOS ONE Assessment of Red Sea temperatures in CMIP5 models for present and future climate

PLOS ONE | https://doi.org/10.1371/journal.pone.0255505 July 30, 2021 26 / 26